Christopher Nobes a a Royal Holloway, University of London, Egham, TW20 0EX, UK. Published online: 14 Apr 2013.

Transcription

1 This article was downloaded by: [Sistema Integrado de Bibliotecas USP] On: 19 December 2014, At: 15:40 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Accounting and Business Research Publication details, including instructions for authors and subscription information: The continued survival of international differences under IFRS Christopher Nobes a a Royal Holloway, University of London, Egham, TW20 0EX, UK Published online: 14 Apr To cite this article: Christopher Nobes (2013) The continued survival of international differences under IFRS, Accounting and Business Research, 43:2, , DOI: / To link to this article: PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content ) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at

2 Accounting and Business Research, 2013 Vol. 43, No. 2, , The continued survival of international differences under IFRS CHRISTOPHER NOBES Royal Holloway, University of London, Egham TW20 0EX, UK The claimed starting point for much recent literature is that International Financial Reporting Standards (IFRS) have been very widely adopted. That is somewhere between an error and a misleading simplification. This paper begins by providing an antidote by analysing the degree to which IFRS have not been adopted in the jurisdictions containing the world s largest 16 stock markets. This might help researchers with their institutional settings. The paper then examines several issues which can lead to international differences in IFRS practice, starting with language and enforcement, but focusing mainly on policy options. Previously published lists of these are up-dated, the extensive recent literature on IFRS policy choice and policy change is synthesised, and new data are provided. Finally, researchers are offered some lessons from the past and some directions for the future. Keywords: accounting policy choice; IFRS; international differences; adoption; policy options; classification 1. Introduction It is commonplace that International Financial Reporting Standards (IFRS) have spread to over 100 countries (e.g. SEC 2008, IASB 2012, p. 11). Examples of reports about this process are as follows: IFRS is the official reporting standard which was recently adopted by over 100 countries. (Benzacar 2008, p. 26) To date, more than 12,000 companies in over 100 countries have adopted IFRS. (Interfacing 2012) The global rollout of International Financial Reporting Standards is gaining momentum, with more than 100 countries now using IFRS and all of the world s major countries anticipated to be on board within the next few years. (BDO 2012) The number of countries requiring International Financial Reporting Standards (IFRS) for public companies has grown from a relative handful to over 100. (Pacter 2012) Approximately 120 nations and reporting jurisdictions permit or require IFRS for domestic listed companies, although approximately 90 countries have fully conformed with IFRS as promulgated by the IASB and include a statement acknowledging such conformity in audit reports. (AICPA 2013) chris.nobes@rhul.ac.uk # 2013 Taylor & Francis

3 84 C. Nobes This paper is a reminder that such reports should not be taken to imply the imminent death of international accounting differences. The paper results from a commission by the editor of this journal for an up-dating and expansion of Nobes (2006) which the editor had noticed was widely cited by researchers. 1 Many have used the logic of that paper as background for hypotheses on the existence or effects of international differences; and several authors have referenced the paper as the specific starting point for empirical research on such issues as comparability (e.g. Cairns et al. 2011, p. 3, Taplin 2011, p. 384). Other researchers have used the paper s lists of IFRS options as dependent variables. 2 The paper suggested eight ways in which national aspects of reporting might survive the arrival of IFRS. These are re-examined and up-dated here. The first contribution of this paper is to present data on the variety in national implementations of IFRS, stressing where implementation is partial or non-mandatory, in order to act as an antidote to the lack of caution in some of the quotations above. Therefore, my inaugural proposition will be: ask not what a company in a country can do; ask what the company is required to do. More specifically, ask not which country allows some version of IFRS for some purpose; ask which country requires IFRS as issued by the International Accounting Standards Board (IASB) (IASB-IFRS) for all purposes. Related to this, there are two other regulatory issues which differ by jurisdiction: translations of IFRS and enforcement of IFRS. This paper s second contribution is to summarise and up-date the discussion of these. There is a further, quite different, way in which international accounting differences have survived the arrival of IFRS. That is, even among companies which are fully compliant with IASB-IFRS, there is scope for, and evidence of, a variety of accounting practices, which are associated with such factors as country, industry and size. Nobes (2006) divided the scope of different IFRS practices into five aspects of the content of IFRS, such as overt options and covert options. The next contributions of this paper are to up-date and go beyond the earlier discussion of these, in two ways: (i) by taking account of the many changes to the content of IFRS from 2005 to 2012, and (ii) by summarising and adding to the empirical findings based on IFRS annual reports of 2005 onwards. One example of new data relates to the first set of Canadian IFRS reports, which confirm previous findings of the persistence of pre-ifrs national practices. In its next four sections, this paper examines the above four sets of issues (different national implementations of IFRS; language and enforcement; the scope for variations in IFRS practice; and the empirical search for those variations). Relevant literature is included section by section. The author apologises to readers for the amount of self-citation, but that is inherent in the nature of this commission. Section 6 concludes, and also asks whether any other set of standards (particularly US generally accepted accounting principles (GAAP)) would have faced the same challenges if adopted widely. When examining national implementations (in Sections 2 and 3 below), reference is made to the requirements for financial statements relating to accounting periods ending on 31 December 2012, unless otherwise stated. For variations allowed within IFRS (in Section 4), it is possible 3 to go one year further: to the requirements in force for 31 December 2013 year ends. 2. National implementations of IFRS As noted above, the literature (academic, professional and journalistic) contains many references to widespread or almost universal adoption of IFRS. However, four caveats should be entered. First, in several major capital markets, IFRS are either not required for any purpose (e.g. in Japan or Switzerland) or not even allowed for domestic reporters (e.g. for registrants of the Securities and Exchange Commission (SEC) in the USA). Second, even in jurisdictions where IFRS are required or allowed, this is often restricted in some way (e.g. to listed companies or to

4 Accounting and Business Research 85 consolidated statements). I am not suggesting that these restrictions on IFRS are inappropriate, merely that they cause an unadorned statement about adoption to be misleading. In the quotations at the start of this paper, Pacter (2012) carefully refers to public companies (which is a very small percentage of companies, although economically important) whereas the other authors do not. Third, for many of the more than 100 countries, it is not IASB-IFRS that are required or allowed (for some purposes) but national (e.g. Australian) or regional (e.g. EU) versions of IFRS. Fourth, in many countries where IFRS adoption is alleged to have taken place (e.g. in several EU countries), companies or auditors do not refer to compliance with IASB-IFRS. Whether companies are complying with IASB-IFRS is a different issue from whether IASB- IFRS are required or whether they are referred to. I deal with those issues separately below, but it should be noted immediately that some versions of IFRS do not necessarily imply (and some do not even allow) compliance with IASB-IFRS. Zeff and Nobes (2010) investigate a number of national implementations of IFRS. They ask, for example: Has Australia adopted IFRS?, concluding that there is no simple answer. Strictly speaking, Australia has not adopted IASB-IFRS because, in Australian IFRS (AIFRS): (i) early adoption 4 as allowed by most new or amended IASB standards is not allowed, (ii) some disclosure requirements have been added, (iii) many paragraphs have been amended to extend application to not-forprofit entities, and (iv) references are added. The lack of permission to adopt early is a major issue: for example, for accounting periods ending on 30 June 2012 or 30 June 2013, a large number of IFRS were allowed in South Africa or Switzerland but not allowed 5 in Australia. This included, inter alia, IFRS 9 13 and important amendments to IAS 19. This material occupied over 500 pages of IFRS, including vital issues such as the definition of a subsidiary. On the other hand, those for-profit entities which comply with AIFRS should automatically therefore be complying with IASB-IFRS. Thus, in terms of comparability, the result is the same as if IASB-IFRS had been required in Australia. More recently, there has been another relevant development in Australia, in that companies without public accountability 6 are allowed to follow a special version of AIFRS which contains reduced disclosure requirements, so is clearly not IASB-IFRS. EU-adopted IFRS are also not IASB-IFRS in that (i) they have a more permissive version of IAS 39 (Whittington 2005), and (ii) early adoption is not allowed until EU endorsement. 7 Despite this, at least up to the time of writing, EU companies can easily arrange 8 to comply with IASB- IFRS, and most are probably 9 doing so even though few acknowledge that they are (Nobes and Zeff 2008). By contrast, Chinese or Venezuelan versions of IFRS would probably not normally allow compliance with IASB-IFRS. Zeff and Nobes (2010) conclude that very few countries with significant stock markets require direct compliance with IASB-IFRS. They mention Israel and South Africa but, even in those countries, compliance is only required for listed companies. The impression of widespread IFRS adoption can be created or dispelled by careful choice of the questions being asked. Table 1 asks questions, at the head of its columns, designed as an antidote to the conventional wisdom outlined in Section 1. The table reports on IFRS implementations in the 16 jurisdictions which have the world s largest stock markets. 10 The table deals with the regulated reporting of domestic companies. It deals only with requirements not with options (except in column 8). Therefore, various permissions to use IFRS are not mentioned, e.g. (i) permission in the USA for foreign SEC registrants or (ii) permission in Japan for Japanese companies. A table which tried to deal with permissions would be very complex, and would not be relevant for the purpose here. Column 2 of Table 1 records that IASB-IFRS are not required in any of the 16 countries for the reporting of all types of entities which are regulated. There would be a more efficient way of conveying this information, but that display suits the purpose outlined in the previous paragraph. Column 3 reports on countries where IASB-IFRS are required for at least some types of reporting

5 Table 1. IFRS implementations for regulated reporting by domestic companies, 31 December 2012 year ends Jurisdiction Is IASB- IFRS required for all regulated reporting? Is IASB-IFRS required for consolidated statements of all listed companies? Is a version of IFRS which is intended to ensure compliance with IASB-IFRS required for all regulated reporting? Is a version of IFRS which is intended to ensure compliance with IASB-IFRS required for consolidated statements of all listed companies? Is a version of IFRS which allows compliance with IASB-IFRS required for all regulated reporting? Is a version of IFRS which allows compliance with IASB-IFRS required for consolidated statements of all listed companies? Australia N N N Y a (2005) N Y N (A) Brazil N N N Y b (2010) N Y N Canada N N c N N N N N d (A) China N N N N N N N France N N N N N Y (2005) N Germany N N N N N Y (2007) e N f Hong N N N N N N N Kong g India N N N N h N N N Japan N N i N N N N N Russia N Y (2012) j N Y N Y N South N Y (2005) N Y N Y N (A) Africa South N N N Y k (2011) N Y N (A) Korea Spain N N N N N Y (2005) N Switzerland N N l N N N N N UK N N N N N Y (2005) N (A) USA N N N N N N N m Is a version of IFRS which allows compliance with IASB- IFRS required (or allowed; A) for all regulated unconsolidated reporting? 86 C. Nobes

6 Notes: Y ¼ Yes. Y ¼ Yes, as follows from a column to the left. N ¼ No. A ¼ Allowed. a Many additions to IASB-IFRS. See text for explanation. b Some IFRS options have been deleted, and some disclosures added. c Rate-regulated companies are exempted. Companies which are also listed in the USA are allowed to use US GAAP. d Private entities are allowed to use the Accounting Standard for Private Enterprises. e US GAAP was allowed for certain companies until f No, unless statements prepared according to the Commercial Code are also prepared. g Hong Kong Financial Reporting Standards have been converged with IFRS but differences remain, primarily in transitional provisions. h The Deloitte and PwC websites concur that the status of IFRS in India is sufficiently unclear that N is the answer. i Japan allows IASB-IFRS for the consolidated statements of listed companies. j The Deloitte and PwC websites concur that IFRS (with an IASB-approved translation into Russian) is required for listed companies. k IASB-IFRS in Korean translation, and without early adoptions permitted. l IFRS or US GAAP are allowed. m Unconsolidated reporting is not generally required in the USA. To the extent that it is required, US GAAP is mandated. Sources: and (both accessed on 29 March 2012). Accounting and Business Research 87

7 88 C. Nobes (e.g. the consolidated statements of all listed companies). Only Russia and South Africa are recorded as Yes. I treat the Russian translation of IFRS as IASB-IFRS because it has been approved by the IASB. Whether Russian companies are fully complying with this recent requirement to use IFRS is a different (empirical) matter. Canada is not included in column 3 because it does not require IFRS for Canadian companies which are rate-regulated or are also listed in the USA; for example, of the 60 members of the main stock market index (TMX S&P), only 49 companies used IFRS during Columns 4 and 5 concern versions of IFRS which are intended by regulators to lead to compliance with IASB-IFRS. No extra countries appear with a Yes in column 4. Those three countries appearing for the first time in column 5 require, for some reporting, a version of IFRS that is not as issued by the IASB because, for example, it does not allow early adoptions of new or amended standards. Columns 6 and 7 concern versions of IFRS which currently allow companies to achieve compliance with IASB-IFRS. No extra countries appear in column 6, but column 7 contains the EU countries. Column 8 concerns unconsolidated statements. No version of IFRS is required for this purpose for all regulated companies in any of the jurisdictions. It is not generally the topic of this paper to investigate the many jurisdictions where versions of IFRS are allowed. However, relaxing the strict approach, column 8 also records any cases where a version of IFRS is allowed for unconsolidated reporting. The UK is the only EU country in Table 1 which is recorded as (A) because the others require national GAAP to be used for the unconsolidated statements from which tax and distribution are calculated. In summary, because of the way of asking the questions, there is little Yes in Table 1. Only 9 of the 16 countries have any Yes at all. It would be easy to reduce the Yes to a minority by deleting the three countries with the smallest of the 16 stock markets (South Korea, South Africa and Spain). As we move to the right across Table 1, a date is inserted for the first appearance of Yes for any country. The date refers to the earliest accounting period ending on 31 December for which the Yes applies. As may be seen, there are no such implementations before 2005, and three out of nine implementations are recent, relating to 2010 or later. Researchers might find it valuable to consider Table 1 when assessing the reliability of reports about IFRS adoption. For example, Francis et al. (2008) present a table showing that most unlisted companies in a majority of 56 countries (including many EU countries) had adopted 12 IFRS by 1999/2000. This cannot have been the case (see columns 6 or 8 of Table 1) if adopted means compliance, as it generally does in the empirical and other literature (e.g. Ashbaugh 2001, Ashbaugh and Pincus 2001, Leuz 2003, Jermakowicz 2004, Cuijpers and Buijink 2005, Gassen and Sellhorn 2006, Hoogendoorn 2006, Christensen et al. 2007, Barth et al. 2008). Country of domicile is by no means a complete predictor of which GAAP will be adopted by a company, but it does make some answers unlikely or (without dual reporting) legally impossible. Research work on national implementations has taken many different directions. For example, Leuz (2010) seeks to explain clusters of countries with similar IFRS implementations; Goncharov et al. (2009) examine the economic functions of the dual (IFRS/national) accounting system in Germany. Nothing in this paper should be read as implying that the IASC and IASB have not succeeded in reducing international differences in financial reporting, and therefore in increasing comparability. The consolidated statements of thousands of listed companies are complying with IASB-IFRS, even if many (e.g. nearly all EU companies) do not acknowledge this. It is clear, for example, that German and UK listed companies now publish consolidated statements which are much more similar than before IFRS was adopted, and that both German 13 and UK 14 national rules have been changed towards IFRS.

8 Accounting and Business Research Language and enforcement 3.1 Introduction As discussed in Section 2, the jurisdiction in which an entity operates will determine the version of IFRS with which the entity is required or allowed to comply. However, there are other aspects of nationality that will also affect IFRS practices. Ball (2006) and Nobes (2006) explain how, even if all entities are apparently complying with IFRS, the incentives of preparers and enforcers remain primarily local (Ball 2006, p. 15). Wysocki (2011) shows how the determinants and outcomes of IFRS are intertwined with non-accounting institutions. Section 4 of this paper focuses on issues of judgement and choice with respect to policies on recognition, measurement and presentation. Before that, this section looks at two other national issues: language and enforcement. From here on, a large number of accounting standards are referred to. In order to avoid clutter, they are referred to by their numbers only, but the full title of each of these standards is shown in the Appendix to this paper. The issues of language and enforcement can be introduced with the aid of a typical 15 audit report on IFRS financial statements in the EU. This one relates to a German company s 16 consolidated statements, as follows: We conducted our audit in accordance with 317 HGB and German generally accepted standards for the audit...promulgated by the Institut der Wirtschaftsprüfer... According to 322 Abs. 3 Satz 1 HGB, we state that our audit...has not led to any reservations. In our opinion,...the consolidated statements comply...with IFRSs, as adopted by the EU, and the additional requirements of German commercial law pursuant to 315a Abs. 1 HGB and give a true and fair view...in accordance with these requirements. The auditors reveal (or imply) at least eight national (or EU) aspects of the reporting: (i) the audit is required by German law and its quality would be assessed under German law, (ii) the auditors use German professional audit standards (although these are now closely based on international standards), (iii) the auditors comply with particular German legal requirements for what must be stated, (iv) the accounting standards are IFRS, as adopted by the E.U., which (as explained above) is not the same thing as IASB-IFRS even though the company might be complying with IASB-IFRS, (v) because the reporting is under German law, it is the German language version of EU-adopted IFRS that applies legally, (vi) German law contains extra accounting requirements, (vii) the auditors apparently 17 report on true and fair rather than on the fair required by IAS 1 (see Section 3.2), and (viii) German law contains particular references to true and fair which weaken the override (see Section 3.2). The audit points contained here could affect the quality of reporting because they might affect compliance (see Section 3.3). 3.2 Language Evans (2004) examines language difficulties in detail in the context of accounting. She points out the inevitability that some meaning will be lost in translation. Zeff (2007) includes problems of translation and terminology in his survey of obstacles to global comparability of financial reporting. The translations of true and fair view and fair presentation were investigated in detail for several languages by Nobes (2009, pp ). The various official translations have rendered true and fair and fair identically, although the German translations use different words for view and presentation. 18 When translating German documents the other way, into English, the translators have to remember to use true and fair view in a legal context (e.g. for the audit report quoted in Section 3.1) but fair presentation in the context of IAS 1.

9 90 C. Nobes German law does not include the true and fair override despite its being required by the EU s Fourth Directive on company law (Nobes 1993). However, the German version of EU-adopted IFRS does include the fair presentation override of IAS 1. Nobes (2009) investigates the use of the override in the context of IFRS, including the famous case of Société Générale in which, on its 2007 financial statements, the company received a clean audit report from two Big-4 firms after explaining that it had departed from IASs 1, 10 and 37. To the extent that this would have been less likely to happen in Germany or the UK, cultural rather than linguistic explanations might be needed (Standish 2003, Burlaud and Colasse 2011, Danjou and Walton 2012). More simply, there are some clear errors of translation. To give a German example, the discount rate required by IAS 19 (para. 83) for pension obligations is that on corporate bonds. However, the official EU German text renders this as Industrieanleihen. Such industrial bonds are a sub-set of corporate bonds, perhaps with different, or differently volatile, average interest rates. The IASB s approved German translation (but not the EU s) was corrected to Unternehmensanleihen in Baskerville and Evans (2011) examine the complex issues to be considered when translating IFRS. Dahlgren and Nilsson (2012) give many detailed examples of translating IFRS into Swedish, noting several straight-forward errors. However, they make the more general point that, because conceptual structures in different languages do not match perfectly, some accounting concepts are simply not translatable (p. 57). 3.3 Monitoring and enforcement The quality of audit and of other monitoring of compliance is a national issue, although there is some co-ordination and exchange of information on enforcement at the EU level through the European Securities and Markets Authority. Efforts have been made to avoid national or European interpretations of IFRS, but that topic takes us into the following section of this paper, which is concerned with IFRS policies. Ball et al. (2000) suggest that the strength of enforcement is related to a country s position in the common/code law classification. Brown and Tarca (2005) examine the varying strength of the enforcement agencies of some countries. If auditors and companies are weakly monitored or subjected to little enforcement in a particular country, non-compliance with IFRS is likely to occur. Non-compliance is one cause of different practices even within jurisdictions which apparently require IFRS. At one stage, for example, it was suggested that non-compliance with IFRS disclosure requirements by German companies was substantial (Street and Bryant 2000, Street and Gray 2001, Glaum and Street 2003). German compliance might now have greatly improved (Berger 2010, Meyer 2011, Hitz et al. 2012), but doubts remain about how fully companies in other countries, such as Russia (Krylova 2003), will comply with IFRS. A pan-european survey is provided by Glaum et al. (2013), who reveal differential compliance with IFRS disclosures related to factors of both country and company (e.g. listing status and auditor). 4. IFRS policies of recognition, measurement and presentation 4.1 Comparability In this section, it is assumed that all the companies being considered are properly complying with IASB-IFRS. That is, we set aside any differences in practice caused by different national versions of IFRS, by translations of IFRS or by non-compliance with IFRS. IFRS (even the original English version of IASB-IFRS) contain various types of scope for varied practice. One context in which to think about varied practice, and why it matters, is comparability. This quality is central to the objectives of IFRS financial reporting (IASC 1989, paras.

10 Accounting and Business Research , IASB 2010, paras. QC 20 25). In Nobes (2006) and several subsequent papers, it was assumed that comparability is undermined by any differences in IFRS practices which are associated with the nationality 19 of a company. However, Jaafar and McLeay (2007) and Nobes and Stadler (2012) point out that apparently country-driven differences might be connected to internationally different mixes of industries, and that some accounting policies might be industryspecific. For example, the use of weighted average inventory measurement (as opposed to first-in first-out (FIFO)) might be common among a sample of German companies because it is appropriate for a particular industry whose members are numerous in the sample. If that industry is less important in other countries, then the resulting international difference in inventory measurement does not necessarily suggest a lack of comparability. Apart from the issues discussed in Sections 2 and 3 above, the various types of scope for varied IFRS practice identified by Nobes (2006) were: (i) overt options, (ii) covert options or vague criteria, (iii) gaps in the requirements, (iv) estimations and (v) first-time adoption issues. Later in this section, the examples of types (i) and (ii) will be up-dated and expanded, focussing on what might be useful for researchers. Before that, let us deal briefly with the other three types. 4.2 Gaps in IFRS, measurement estimations and first-time adoption Gaps in any GAAP are, of course, inevitable. IAS 8 deals with the selection of accounting policies when IFRS do not deal with an issue. This leaves much room for judgement and therefore for different solutions. For example, at the time of writing, IFRS 4 and 6 do not provide requirements on several major issues of accounting for insurance contracts and for oil and gas exploration, respectively. Indeed, those standards allow more flexibility than IAS 8 normally does (see Table 2). The gaps in IFRS were mentioned by the SEC (2011) as one reason for wariness about implementing IFRS in the USA. US GAAP does have requirements in the above areas. Nobes (2006, Table 4) recorded a number of areas in IFRS which require measurement estimations, such as assessing an asset s useful life, its pattern of wearing out and its residual value (in order to calculate depreciation). The examples in that table all remain in IFRS in An important recent addition is several estimations needed for the disclosures on financial instruments required by IFRS 7, such as exposures to risks (paras. 33 and 36). These remarks are not a particular criticism of IFRS, because any GAAP would require many of these measurement estimations. However, they are part of the scope for international variations in practice. First-time adoption issues relate to the many choices in IFRS 1. For example, in its pre-ifrs accounting, a company might have treated goodwill in one of several ways, such as: (i) held at zero, with a corresponding deduction from reserves, (ii) held on a cost basis, amortised over 5, 20 or 40 years, or (iii) held on a cost basis but checked for impairment every year. Although IFRS 3 requires method (iii), a company is allowed, on first-time adoption of IFRS, to retain the goodwill amount resulting from any other method in its first opening IFRS balance sheet. The large differences between goodwill numbers prepared under the various bases can last for many years 20 after first IFRS adoption. 4.3 Overt options An IFRS option is called overt here if it is plainly specified as a choice 21 within a standard. By contrast, covert options (discussed in the next sub-section) exist where no choice is explicitly offered but where the degree of judgement involved might allow scope for the preferences of the preparers of financial statements.

11 92 C. Nobes Table 2. Examples of overt options in IFRS in 2013 a. IAS 1 No format requirements for balance sheets or income statements (paras. 54, 82) IAS 1 Choice of by-nature or by-function income statement (para. 99) IAS 1 Permission to show comprehensive income in two statements (para. 81) IAS 2 Permission to measure certain types of inventories at net realisable value, with gains and losses to profit and loss (para. 3) IAS 2 FIFO or weighted average for the determination of the cost of inventories (para. 25) IAS 7 Choice of direct or indirect calculation of operating cash flows (para. 18) IAS 7 Net basis allowed for some cash flows (para. 21) IAS 7 Choice of classification for interest and dividend flows (para. 31) IAS 8 When developing an accounting policy, an entity may also consider the rules of certain other standard-setters (para. 12) IAS 16 Cost or fair value measurement for classes of property, plant and equipment (para. 29) IAS 19 Inclusion of defined benefit interest cost as finance expense or as operating expense (BC para. 202) IAS 20 Asset grants can be presented as a deduction from the asset or as deferred income (para. 24) IAS 21 Choice of presentation currency (para. 38) IAS 27 In separate statements, investments in subsidiaries, joint ventures and associates may be shown at cost or as available-for-sale financial assets (para. 10) IAS 28 In consolidated statements, certain investments in joint ventures and associates may be measured at fair value through profit or loss (para. 18) IAS 38 Cost or fair value measurement for some types of intangible asset (para. 72) IAS 39 Choice of cost basis or marking to market for some designated financial assets and liabilities (para. 9). (Other choices are also available within para. 9) IAS 39 Trade date or settlement date accounting for regular way transactions in financial assets (para. 38) IAS 39 Permission to re-classify financial assets out of trading in rare circumstances (para. 50 (c)) IAS 39 Hedge accounting based on the designation of a hedging relationship (para. 71) IAS 40 Permission to classify a property held under an operating lease as an investment property (para. 6) IAS 40 Entity-wide choice of cost or fair value for measurement of investment property (para. 30) IFRS 3 For measurement of a non-controlling interest, a choice of fair value or the share of the acquiree s net assets (para. 19) IFRS 4 Permission for insurers not to separate certain embedded derivatives (para. 8) IFRS 4 Permission to unbundle components of an insurance contract (para. 10) IFRS 4 Permission to depart from IAS 8 when developing certain accounting policies for insurance contracts (para. 14) IFRS 6 Permission to depart from aspects of IAS 8 when developing accounting policies for exploration assets (para. 6) IFRS 6 Cost or fair value measurement for exploration assets (para. 12) IFRS 9 Trade date or settlement date accounting for regular way transactions in financial assets (para ) IFRS 9 Option to designate certain financial assets and liabilities at fair value through profit or loss (paras and 4.2.2) IFRS 9 Option to present fair value changes on non-trading financial assets in OCI (para ) a For accounting periods beginning on or after 1 January Paragraph numbers as at 30 April Table 2 expands and updates the list of overt options found in Nobes (2006). Researchers might wish to divide these into presentation and recognition/measurement topics. However, this is sometimes ambiguous. For example, it is clear that the IAS 1 options in Table 2 relate to presentation and that the IAS 16 or IAS 40 options are measurement. However, the IAS 20 option which is expressed in the standard as an issue of presentation affects the measurement of assets. The IAS 7 options, again expressed as presentation, change totals such as operating cash flows. IFRS have historically contained more options than most national GAAPs, because of the negotiations involved in setting standards internationally (e.g. Camfferman and Zeff 2006, p. 269). For example, US GAAP contains considerably fewer options, 22 although still allowing LIFO, which has been removed from IFRS. Over time, the IASB has removed options, in

12 Accounting and Business Research 93 order to improve comparability. Researchers would need to check which options were available for which period. Table 3 shows the options which have been removed for various periods after However, the IASB has not been able to avoid 23 adding options, such as that related to IFRS 3 in Table 2. Although all the overt options are clearly visible in the standards, this does not mean that they are all well-suited to research. This is because: (i) some topics occur infrequently (e.g. in Table 2, the first IAS 2 choice, the IAS 8 choice, the IAS 38 choice and the first IAS 40 choice); and (ii) some relate to specific industries which the researcher might not be studying (e.g. the second IAS 39 choice, and the IFRS 4 and IFRS 6 issues). Removing the rare issues leads us to a set of easily researched choices, as set out in Table 4. These have formed the basis of recent empirical research which is discussed in Section 5. Table 4 includes four items (shown with daggers) which are no longer options in 2013 but it excludes some items which have proved difficult to assess in practice Covert options Turning to covert options, Table 5 presents a list which is an expanded and up-dated version of that provided in Nobes (2006, Table 2). An example of an important covert option is the degree to which development costs are capitalised, which depends on the interpretation of several somewhat vague criteria. As with overt options, IFRS contains more of them than some other GAAPs. For example, the one in the previous sentence does not apply generally 25 in US GAAP, and neither do several others. 26 Not only are covert options hidden from view in the standards; the exercise of such options is also generally hidden from view in financial statements. For example, it is usually impossible for outside users of the statements to assess whether or how a company exercised its preferences when capitalising (or not) development costs. One exception to this is the depreciation method, which is often disclosed. IASs 16, 38 and 40 make it clear 27 that selecting the Table 3. Former overt options in IFRS a. Until (and including) 2004 IAS 2 LIFO basis allowed (para. 23) IAS 8 Policy changes and error corrections could be taken through the income statement (para. 38) IAS 21 Certain exchange differences from severe devaluation could be capitalised (para. 21) IAS 27 Option to present minority interests outside of equity (para. 26) Until (and including) 2008 IAS 1 Choice of two types of presentation of OCI (para. 10) IAS 23 Choice of capitalising or expensing interest costs on construction contracts (paras. 7 and 10) Until (and including) 2012 IAS 19 AGL could be taken (a) immediately in full to OCI, (b) immediately in full profit or loss, (c) in full to profit or loss over the remaining useful lives of employees in the plan, (d) in part to profit or loss over that period (corridor method), (e) in full or in part to profit or loss over a shorter period (paras A) IAS 31 In consolidated statements, there was a choice of proportional consolidation or equity accounting for joint venture entities (para. 30) a Paragraph numbers relate to the last versions of the IASs before the removal of the relevant option.

13 94 C. Nobes Table 4. IFRS policy choices that are or were a easily observed. 1 b (a) Balance sheet shows assets ¼ credits (b) Shows net assets 2 b (a) Liquidity decreasing (cash at top) (b) Liquidity increasing 3 b (a) Income statement by function (b) By nature (c) Neither 4 b (a) Equity accounting results included in operating profit (b) Immediately after operating profit (c) After finance 5 a (a) Statement of changes in equity, including dividends and share issues (b) SORIE/OCI, not including them 6 b (a) Direct calculation of operating cash flow (b) Indirect 7 b (a) Interest paid shown as operating cash outflow (b) As financing 8 (a) Only cost for PPE (b) Some fair value 9 (a) Investment property at cost (b) At fair value 10 b (a) Some designation of financial assets at fair value (b) None 11 a (a) Capitalisation of interest on construction (b) Expensing 12 b (a) FIFO for inventory cost (b) Weighted average 13 a (a) Actuarial gains and losses to OCI (b) To income in full (c) Corridor method 14 a (a) Proportional consolidation of joint ventures (b) Equity method a Options removed for 2013 statements or before; see Table 3. b Non-financial companies only. depreciation method is not a policy choice, because it should rely on estimating the pattern of consumption of the asset s benefits. Just as the list of overt options has changed over time, so has the list of covert options. For example, until the 2003 version of IAS 1, entities were required to present any extraordinary items in their income statements, and the determination of whether an item was extraordinary involved judgement. Given that it is much more difficult to research covert options, the treatment of them here is not as detailed as for overt options. 4.5 Scope for country differences The scope for variation in IFRS practices, as examined in this section, could undermine comparability. This would be the case even if the scope were exercised randomly by companies, but it was suggested (e.g. Ball 2006, Nobes 2006) that there will be systematic country-based differences in practice. These might be influenced by the same forces which drove national GAAPs to be different in the first place, such as different tax systems and different prime purposes of accounting, linked to different owners and financiers of companies. The influence can be indirect. For example, tax considerations could influence a company s accounting policies (e.g. choosing

14 Accounting and Business Research 95 Table 5. Examples of covert options or vague criteria in IFRS in 2013 a. IAS 1 Departure from the requirements of IFRS when necessary to comply with the objective of financial statements (para. 19) IAS 1 Determination of whether a liability is current on the basis of expected date of settlement or purpose of holding (para. 60) IAS 8 Determination of materiality for various purposes (para. 5) IAS 11 Use of stage of completion method for construction if contract outcome can be estimated reliably (para. 22) IAS 12 Recognition of deferred tax asset for a loss carryforward only if future taxable profit is probable (para. 34) IAS 12 Recognition of deferred tax liability on unremitted profits from subsidiaries only if dividends are probable in the foreseeable future (para. 39) IAS 16 Depreciation method (paras ) IAS 17 Lease classification based on substantially all the risks and rewards with no numerical criteria (para. 8) IAS 18 Use of stage of completion method for services if contract outcome can be estimated reliably (para. 38) IAS 21 Determination of functional currency based on a mixture of criteria (paras. 9 12) IAS 28 The identification of an associate on the basis of significant influence (para. 2) IAS 32 Offsetting of financial asset and liability when there is intent to settle on a net basis (para. 42) IAS 36 Identification of an indication of impairment based on a mixture of criteria (paras ) IAS 37 Recognition of a provision based on probability of outflow of resources (para. 14) IAS 38 Capitalisation of development costs when all of various criteria are met (para. 57) IAS 38 Amortisation of intangible assets only if useful life is assessed as finite (para. 88) IAS 39 Classification of financial assets based on intention to trade, hold-to-maturity or neither (para. 9) IAS 39 Use of cost basis where equity instruments cannot be measured reliably (para. 46) IAS 39 Estimation of hedge effectiveness as a condition for use of hedge accounting (para. 88) IAS 40 Use of cost basis, despite entity-wide choice of fair value, for an investment property whose fair value cannot be measured reliably (para. 53) IAS 41 Use of cost basis for a biological asset whose fair value cannot be measured reliably (para. 30) IFRS 3 Identifying the acquirer in a business combination presented as a merger of equals (para. 20) IFRS 5 Treatment of assets as held-for-sale if expected to be sold within one year (para. 8) IFRS 8 Identification of operating segments based on whether results are reviewed internally (para. 5) IFRS 9 Classification of financial assets based on the entity s business model (para ) IFRS 10 Identification of a subsidiary based on assessing power over an investee (para. 17) IFRS 11 Identification of a joint venture based on assessing relevant activities (para. 7) a For accounting periods beginning on or after 1 January Paragraph numbers as at 30 April weighted average inventory costing) or estimations (e.g. maximising the size of impairments or provisions) in unconsolidated statements under a national GAAP; and these might flow through to IFRS consolidated statements. Even if such forces are no longer relevant, companies might prefer to continue with previous practices for administrative ease or to provide continuity for users. 5. Empirical studies of variations in IFRS practice 5.1 Introduction There have been several studies which collect data on IFRS practices, mostly relating to 2005 financial statements or later. Some of these studies record the overt IFRS options chosen by companies on some topics (e.g. KPMG and von Keitz 2006, ICAEW 2007, European Commission 2008). Those studies mix large and small companies, and have small samples per country. They generally do not report data by country. However, several empirical studies have systematically collected data for a formal sample of companies on a country-by-country basis, using lists

15 96 C. Nobes of overt options for which company choice is easily observable, similar to that in Table 4. These studies are now discussed in the next five sub-sections, followed by a sub-section containing new analogous data on Canada. 5.2 IFRS policy choices in 2005/2006: national patterns Kvaal and Nobes (2010, hereafter K&N) examine the 2005/2006 practices of 232 large listed companies from the five IFRS-using countries with the largest stock markets: Australia, France, Germany, Spain and the UK. As in all the papers discussed in this section, foreign companies 28 are excluded. For 16 overt IFRS options, 29 K&N s general hypothesis (based on the reasoning outlined in Sub-section 4.5) is not exactly that country will determine policy but that any particular company will continue to follow its own individual pre-ifrs national practice on each particular optional topic. This motivated K&N to collect data on the pre-ifrs policies of their companies for the topics without pre-ifrs national requirements. K&N then hand-picked the data on the IFRS policy choices made by their companies. K&N find very strong statistical evidence that these choices follow pre-ifrs practices and therefore that there are national patterns of IFRS practice, which undermine comparability. K&N note that some of their 16 topics are of little importance in this context but that they all bolster the findings on the importance of nationality in explaining policy choice. This increases the probability that there are national differences in other important areas which cannot be easily observed, such as the covert options related to impairment or to the capitalisation of development costs (see Table 5). K&N do not examine the potential importance of a company s sector in explaining IFRS policy choice, although they exclude financial companies on topics for which a sector-specific choice is likely. Table 4 shows these topics with asterisks. K&N disclose the sectoral mix of their sample, which is approximately 30 as in Table 6 of this paper. They note that (apart from financial companies) there is a broad sectoral spread for all their five countries. Sector is discussed further below. Studies of particular IFRS policy areas also uncover a national effect. Morais (2008) and Fasshauer et al. (2008) look at the treatment of actuarial gains and losses (AGL) in the 2005 IFRS reports of European companies. Both studies find an influence of nationality; for example, that UK and Irish companies use different methods from continental companies. Cairns et al. (2011) focus on the use of fair value (including options to use it) in Australia and the UK around the time of transition to IFRS. They did not find great difference between the Table 6. Country and sector a distribution of companies, 2008/2009 data (Canada, 2011). Australia Canada UK Germany France Spain Netherlands Italy Sweden 0 Oil and gas Basic materials Industrials Consumer goods Health care Consumer services Telecommunications Utilities Financials Technology Total companies a Sectors according to Industry Classification Benchmark.

16 Accounting and Business Research 97 two countries, and this is not surprising given the similarities of background. Verriest et al. (2013) examine first-time adoption of IFRS, and discover national variations, possibly due to varying quality of corporate governance. 5.3 IFRS policy changes from 2005/2006 to 2008/2009 For most of the 232 companies in K&N s sample, the financial statements examined were the first prepared under IFRS. This raises the question whether the influence of pre-ifrs practices (and the resulting patterns of national IFRS practice) was mostly a feature of transition to IFRS and might therefore dissipate over time. In order to investigate that, Kvaal and Nobes (2012) examine the IFRS policy choices of the same companies as in K&N but three years later. They expect the influence of pre-ifrs policies (and therefore nationality) to continue, not least because IAS 8 discourages 31 policy change. Kvaal and Nobes (2012) examine the 126 policy changes over their three-year period. They discover little change among Australian and UK companies but significantly greater change by French and Spanish companies. Remarkably, the French and Spanish companies changed their policies more after transition (despite the constraints of IAS 8) than they had done at transition (when IFRS 1 imposes no constraints). For Australian and UK companies, the reverse was the case. Kvaal and Nobes (2012) ask what might explain this. They suggest that French and Spanish companies had been the furthest away from the culture, ethos and requirements of the IASB, and so were least aware of the possibilities for change at transition. K&N test this by recording, for each post-transition policy change, whether the change was towards or away from the pre- IFRS requirement under national GAAP. They find that most of the few Australian and UK changes were towards the pre-ifrs national requirement, suggesting that the companies were not learning from foreign companies, but were perhaps interested in comparability among their national peers. However, the many French and Spanish post-transition changes were mostly away from the pre-ifrs national norms, suggesting some learning, perhaps from the practices of companies listed on the London stock exchange as the largest in the world to be based on reporting under IFRS. As part of this exercise, K&N checked whether any amendments or proposed amendments to standards could explain the policy changes. There were no relevant amendments to IFRS requirements in the period and, on average, companies took no policy action in response to amendments proposed by the IASB. For example, there was no net movement away from the proportional consolidation of joint ventures after the IASB announced plans to abolish the method; and there was no increase in the move towards taking AGL to other comprehensive income (OCI) after the IASB announced plans to require this. Haller and Wehrfritz (2012) also use Nobes (2006) as a starting point. They examine the IFRS policy changes of German and UK companies from 2005 to Like Kvaal and Nobes (2012), they find few policy changes among companies of those countries. Australia presents a special case because the initial Australian implementation of IFRS had deleted certain non-australian options, such as the indirect method of calculating operating cash flows (topic 6 in Table 4). However, for 2007 onwards, the options were reinserted. Bond et al. (2012) find that less than 1% of all Australian listed companies had changed from the direct to the indirect method by This confirms the strong influence of pre-ifrs practices. 5.4 Country, sector, firm and topic influences on IFRS policies As mentioned in Section 4.1, Jaafar and McLeay (2007) investigated whether sector-specific policy choices lay behind some of the international differences. They found only a small effect